Kwik-Fit Group Limited & Ors v The Commissioners for HMRC

Before the FTT, Kwik-fit submitted that the case-law (IRC v Parker [1966] AC 141 and IRC v Sema GroupPension Scheme Trustees [2002] EWCA Civ 1857) required an exercise of comparing the actual tax consequences with those which might have transpired to see if the actual tax outcome was more favourable than the comparator. The FTT disagreed that such a comparator transaction was required, referring to Sema Group (which we discuss in more detail below) where it was said “relief” covered situations where the taxpayer’s liability is reduced, leaving a smaller sum to be paid, and that a tax advantage presupposed that a better position vis-à-vis the Revenue had been achieved (i.e. a better position than if such relief had not been available). The FTT agreed with HMRC that the use by Speedy of its NTLRDs to offset against its interest income was a “relief from tax” “as without those NTDs being available, Speedy 1 would have been required to pay tax on its net interest income”([77]).

Mr Ghosh KC, for the appellants accepts the FTT correctly identified the test for a “tax advantage” (and was also correct that the legislation did not require postulation of a comparator transaction). But, he argues, the FTT misapplied the test, because as a matter of principle, use of existing tax losses to shelter newly acquired income could not comprise a tax advantage. He emphasised it was wrong to regard the functioning of a provision, doing exactly what it was intended to do, as a tax advantage. Utilisation of losses against profits was not a relief. Speedy already had the relief of losses. Its tax liability and its position vis-à-vis HMRC were unchanged as before it had losses, and no charge to tax, and after (the losses having sheltered the newly acquired income) it still had no charge to tax.

The appellants’ recourse to Speedy’s position vis-à-vis HMRC remaining the same echoes the discussion of the provisions on “tax advantage” in Sema Group. The analysis must however start with the words of actual legislation and the central question, namely whether utilisation of the NTLRDs was a “relief”.

In interpreting the term “relief” it is necessary to look at the term in its statutory context. As Ms Wilson KC, for HMRC, noted, the definition of “tax advantage” in s1139(2) CTA 2010 is comprehensive in the sense that the sub-section refers to a number of stages in the taxing process: the tax, the charge to tax, and its potential assessment or assessment, and in relation to any repayment of tax. As a matter of drafting, the separate reference to “reduction of a charge to tax or an assessment to tax” in subsection (c) suggests that the earlier reference to “relief” in 1139(2)(a) is not necessarily restricted to situations where the ultimate tax chargeable or assessed is altered.

Mr Ghosh argues it cannot be a tax advantage to use the losses up more quickly thereby accelerating an eventual charge to tax. Utilisation of losses is a disadvantage (or the counteraction of the advantage of having losses available that could otherwise be used up).

However, putting the issue in this way does not assist. It relies on looking at the meaning of “tax advantage” in a general sense. It is not necessary to test whether what is sought to be secured is a tax advantage in any generalised sense where Parliament has provided a particular definition, which includes within it securing anything which is a “relief”.

That term, in our judgment, readily encompasses the notion of the tax code allowing less tax to be charged in a context where the provision is doing exactly what it was intended to do, namely providing relief from tax. On a plain reading of the legislation, we see no difficulty in regarding a provision under which a deficit is set off against profits, as a relief from tax. The setting off reduces the amount of profits on which tax is charged thereby reducing the tax.

For the purposes of 1139(2)(a) it is not necessary to determine whether the relief ultimately does result in less tax being payable, but just whether, functionally, the provision operates in that way.

We also cannot see, in any case, how it can be argued, as the appellants do, that use of the losses was not a tax advantage because by using losses there was then less of them to use. It is inherent in the nature of a relief that is dependent on the particular amount, that the more of it that is used, the less of it there will be. The fact the relief is used does not make it any less a relief.

The proposition that utilisation of losses may be regarded as a relief is conceptually no different from the deduction of loan relationship debits from interest income being accepted as a “tax advantage” as it was in HMRC v BlackRock Holdco 5 LLC [2022] UKUT 199 (at [127])and as it is by the appellants (although, as we will come onto in our discussion of 1b, the appellants argue that in their particular factual circumstances the deduction cannot be a relief). In both cases a figure, in relation to which the greater the figure, the greater the tax which is charged, is reduced.

Mr Ghosh sought to contrast the straightforward utilisation of losses with the scheme in Versteegh Ltd and others v HMRC [2014] SFTD 547 (where the scheme was described to achieve a corporation tax deduction in one group company for the costs of an intra-group borrowing, but without any concomitant taxable accrual or receipt in the group company). He also contrasted the facts in Travel Document Service and another v HMRC [2018] EWCA Civ 549 where, as Mr Ghosh put it, “the debit was created out of thin air”. But the question of what constitutes a “tax advantage” is, as Ms Wilson, aptly described it, “neutrally” framed. The term requires a functional analysis of the relevant legislative provision. The nature of transactions being part of a scheme might be relevant though not conclusive to the question of purpose, but it is no part of the definition of “tax advantage”.

Ms Wilson rightly acknowledges that in a given transaction or relationship there could be many such “tax advantages” in operation. That obviously does not mean a corresponding multitude of debits risk being denied under the unallowable purpose regime. It is only if the company has a purpose of securing the tax advantage, and not only that, that the purpose is a main purpose, that the tax advantage will be relevant.

The FTT, in concluding that the use by Speedy of its NTLRDs to offset against its interest income was a “tax advantage”, had earlier referred to Sema Group and the proposition stated there (see below) that the definition of “tax advantage” in similar legislation sought to cover “every situation in which the position of the taxpayer vis-à-vis the Revenue is improved in consequence of the particular transaction or transactions”.

In Sema Group the appellants, who were trustees of an exempt approved pension scheme, received sums in respect of a buy-back of shares by a company the trustees had previously bought shares from. The buy-back was treated as a distribution by the company. The trustees were specifically exempted from income tax on their investment income. That status meant that, under the legislation at the time, they were entitled to a tax credit which corresponded to the rate of advance corporation tax. The Revenue sought to reverse the payment of that tax credit under anti-avoidance provisions which contained a similar definition of “tax advantage” to that which applies here. The Court of Appeal considered the disputed issue of whether the trustees had obtained a “tax advantage” (this was on an obiter basis given its earlier conclusion that the Revenue’s appeal failed because the anti-avoidance legislation did not apply for other reasons). The trustees argued that there was a conceptual difference between exemption (which removed income from charge) and relief (which assumed a charge to tax but which reduced the tax payable). The Court of Appeal rejected that in the following terms:

The Court of Appeal upheld the High Court’s views that the trustees’ exemption was a relief within the section.

Mr Ghosh submits the FTT erred in failing to distinguish Sema Group. Whereas the taxpayer trustees in Sema Group (who were exempt from tax and in a tax neutral position) obtained a tax credit – cash in their hand - which represented a tax advantage, in the appellants’ case the NTLRDs were already there in Speedy. The reorganisation did not improve Speedy’s position as regards HMRC. Before it, Speedy had losses and no charge to tax. After it Speedy had no charge to tax having sheltered the newly acquired income.

We consider this argument misinterprets what the Court of Appeal meant when it referred to the position of the taxpayer being improved or (as it put it at [110]) “that a better position [had] been achieved vis a vis the Revenue”. This formulation was not put forward as a gloss to the statutory words to suggest that a calculation was to be performed comparing the position vis a vis the Revenue before and after the transaction. Rather, it identified a unifying theme or function behind the various different heads of “tax advantage” mentioned in the legislative provision which explained why, because of its effect in improving the taxpayer’s position, the exemption from income tax applicable to the pension trustees in Sema Group could count as a relief.

If, in effect, the appellants’ argument is that the FTT was wrong to consider Speedy’s position vis-à-vis the Revenue improved, then that is clearly wrong for the reasons already explained that there is a tax advantage where the relief is used even if it does not result in a reduction in tax payable by the company. For example, in Speedy’s case there would be no net tax payable in the given years of operation whether the scheme had been adopted or not: a) where the arrangement was implemented there would be no net profits because the NTLRDs were set off against income from interest – so there was no taxable profit for Speedy; or b) if the arrangement had not been implemented there would no income for the NTLRDs to be set off against so the company would simply register losses that would not give rise to any tax charge. Nevertheless, there was a tax advantage in a) because the relief was obtained and used – the losses were set off against profits. Thus, Speedy’s tax liability was less for that year than if it had not claimed the relief and offset losses against profits.

As regards a tax advantage in the form of a relief (under s1139(2)(a) CTA 2010) it was enough that by virtue of the provision in contention (here the offset of a loss against the increased interest income) the taxpayer’s liability was less than it would be without that offsetting against such income.

We disagree therefore that the FTT made any error in its reference to Sema Group.

In support of an argument that the utilisation of losses could not sensibly be seen as securing a tax advantage, Mr Ghosh referred to obiter statements of the High Court in Commissioners of Inland Revenue v Kleinwort Benson Ltd. [1969] 2 Ch 221with regard to similar legislation (s28 Finance Act 1960). There, the taxpayer bank, which dealt with securities on its own account, bought debenture stock which had accrued a large arrears of interest. When the interest arrears were paid, the taxpayer, as it was entitled to, deducted the interest, the interest being income taxed at source which resulted in a loss on the transactions which reduced the taxpayer’s assessable profits. The Revenue sought to exclude the loss under s28 but lost at first-instance. On appeal to the High Court, Cross J dismissed HMRC’s appeal on the basis that certain pre-requisite circumstances referred to in s28 were not fulfilled but went on to give his views on whether the main object of the transaction was gaining a tax advantage:

We do not consider that Kleinwort advances the appellants’ case regarding the meaning of “tax advantage”. Cross J’s views assume that there is a tax advantage and his examples were primarily concerned with whether the taxpayer had the requisite purpose. It is clear, as noted earlier on by Cross J, that the taxpayer had conceded that it had obtained a tax advantage as a result of the transaction so the interpretation of tax advantage was not even potentially in issue. (To the extent Cross J expressed scepticism in setting out his example, that a right to deduct the £20 sum from gross receipts could be described a “tax advantage” then there is no particular indication that each aspect of the more detailed definition of that term was considered and in particular whether the deduction could be termed a “relief”).

Although Mr Ghosh suggested Cross J’s dicta were approved by the Court of Appeal in Sema Group (at [119]) we agree with Ms Wilson that on closer scrutiny that is not correct. The approval was actually of Lightman J’s rejection (at [53]) of the appellant’s submissions on Cross J’s dicta. Lightman J noted the question of purpose was one of fact that could not be interfered with unless the first instance decision makers had misdirected themselves. He noted that Cross J plainly thought they had in Kleinwort. (It is clear from the above extract that Cross J thought that the misdirection was confusing an intention with purpose). In [119], the Court of Appeal went on to note that Cross J’s observations “must be read in the context of the particular transaction which was in issue in that case.” Beyond serving as a reminder not to conflate intention with purpose we do not consider Cross J’s obiter dicta lay down any principle of general application that assists the appellants’ case.

We thus disagree with any argument based on Kleinwort that, as matter of legal principle, a person cannot have a purpose to secure a tax advantage because what they are doing follows inherently from the legislation. Moreover, as Ms Wilson pointed out, Cross J’s sale transaction example can straightforwardly be understood as a situation where it makes no sense to say the buyer’s purpose is to secure the tax deduction; rather the obvious purpose is to turn a profit.

In a similar vein, Mr Ghosh relied on the FTT’s decision in Burlington Loan Management DAC v HMRC [2022] UKFTT 00290 (TC). There a debt was sold to a company whose residence in Ireland meant it did not suffer UK withholding tax on the interest paid to under Article 12.1 of the UK/Ireland Double Taxation Treaty. The Irish company factored this “tax attribute” into the higher price it was prepared to pay the assignor. The issue was the application of Article 12.5 of the Treaty. That denied the UK withholding tax treaty benefit where any person concerned in the transaction had a main purpose of taking advantage of Article 12.1. The FTT rejected HMRC’s case that there was such a main purpose. The UK withholding tax exemption the Irish company enjoyed “was merely part of the scenery – “setting” in which the Irish company made its offer ([174]). In Mr Ghosh’s submission, the proposition to be drawn from the case was that, where the tax attributes of one party to a transaction are a given, it cannot be a main purpose to take advantage of the attributes simply because the parties know that the relief will be enjoyed. Just as the Irish company’s knowledge it had benefit of the treaty did not mean it had a purpose to take advantage of the treaty or give an advantage to the debt assignor, similarly the knowledge of the debtors in this case that Speedy had losses and were willing to pay the arm’s length rate to help Speedy use those losses did not mean the debtors had a purpose to secure a tax advantage.

We reject any submission that one can draw any wider point from the above that there are circumstances where it is not possible, as a matter of legal principle, to conclude there was an unallowable purpose. Putting aside the different legislation and context in which Burlington arose, the conclusion as to lack of purpose turned on the FTT’s detailed analysis of the facts. The FTT went on to find at [185] that as far as the Irish company’s subjective purposes were concerned it was entering into the transaction solely for its own benefit “ to secure a significant profit for itself”. Whether a tax attribute is indeed merely “setting” or “scenery”, as the FTT concluded, will depend on the particular facts of that case. The case, as Ms Wilson pointed out, confirms the uncontroversial proposition (see [84] onwards below) that knowledge of tax consequences does not equate to having a purpose of taking advantage of them.

The appellants’ skeleton argument alleged the FTT wrongly referred to “group benefit” (at [102(5)]) and “group tax advantage” (in terms of tax advantage arising to Speedy being described as the purpose/one of the purposes of the appellants (at [117]) given that the legislation is clear the securing of a tax advantage was for a legal person (s442(5) CTA 2009). Mr Ghosh was right not to press these points in his oral case. The FTT carefully expressed its conclusions on tax advantage (at [77]) in terms of the debits claimed by the appellants and the use by Speedy of its NTLRDs to offset against its interest income. There is no suggestion the FTT misapprehended the relevant requirement to consider the issue tax advantage in relation to a legal person, or that, as was suggested in Mr Ghosh’s reply, it wrongly considered there was a tax advantage by looking at Speedy and the appellant companies together.

The appellants also submitted the FTT erred in referring, at [82], to a tax avoidance purposes being found by reference to securing a tax advantage for the appellant, any of the other appellants, Speedy “…or indeed any other person”. Mr Ghosh argues the advantaged person needed to be identified, otherwise it would be difficult to say whether securing that tax advantage for that person was a main purpose of the loan relationship.

We agree there is no conceptual difficulty with not knowing the identity of the company in respect of whom a tax advantage is sought to be secured for the reasons Ms Wilson outlined in her response. Sections 441 and 442 exist in a context where taxable profits are not assessed until the end of the period. In a group relief situation, a company will not necessarily know at the outset firstly whether it itself will have profits against which the losses may be offset, or if a surrender is on the cards, which company it best makes sense to surrender the loss to.

The FTT’s decision, in any case, shows it was clearly satisfied a tax advantage was secured for identified persons, namely Speedy, and each appellant. The FTT did not rely, nor need to rely on an advantage being secured for unidentified persons, in coming to the view the purpose to secure a tax advantage was a main purpose. So even assuming certainty as to the person’s identity were required it would not make a difference here.

A further submission the appellants make is based on the application of transfer pricing provisions. In brief, these provisions substitute, in certain group contexts, an arm’s length rate of interest (which in the circumstances of this case would correspond to the increased rate of interest that applied post-restructuring of Kwik-Fit Group’s intra-group debt) where the interest was below arm’s length (as it was pre-restructuring of that intra-group debt). It is argued it makes no sense to say that a tax advantage was secured for Speedy when it should be taxed on that basis anyway. The appellants also rely on the transfer pricing provisions, as we will see, in their case that the FTT erred under Issue 2 (just and reasonable apportionment). It is convenient to outline the provisions here.

The transfer pricing provisions are set out in the Taxation (International and Other Provisions) Act 2010 (“TIOPA”) and, in summary, entail comparing the actual provision made between two members with what would have been provided if the two companies were two independent companies. For the purposes of this case it is sufficient to note the following features.

Where the “actual provision” (s147(4)) “confers a potential advantage in relation to United Kingdom taxation (whether or not the same advantage) on each of the affected persons”, under s147(5), “profits and losses of each of the affected persons are to be calculated for tax purposes as if the arm's length provision had been made or imposed instead of the actual provision”.

Section 155 TIOPA explains an actual provision confers a “potential advantage in relation to United Kingdom taxation” on a person if, inter alia, that person’s taxable profits for a chargeable period are smaller in amount as an effect of the provision not being arm’s length.

Where only one of the affected parties is potentially advantaged by the actual provision and the other affected party is within the charge to corporation tax in respect of profits arising from the activities in the course of which the actual provision was made, the other party (referred to as the “disadvantaged person”) may, subject to conditions, make a claim under s174, so that its taxable profits are calculated as if the arm’s length provision had been made instead of the actual provision. This would thus enable (but not oblige) the debtor to make a claim for a deduction corresponding to arm’s length rate of interest.

Section 446 CTA 2009 provides, in summary, that where, under TIOPA amounts are deemed as profits or losses, interest payable or expenses incurred in relation to loan relationships or related transactions, the credits or debits relating to such amounts are brought into account to the same extent as they would be in the case of actual amounts of such profits, losses, interest or expenses.

The unallowable purposes regime therefore acknowledges and makes provisions for TIOPA deemed amounts in that the debits and credits generated by deemed figures are recognised within the unallowable purpose regime as if those debits and credits were generated by actual figures.

The FTT found (at [49]) that the rate to which the interest rates on the loans were amended (LIBOR + 5%) was arm’s length.

As to the appellants’ argument regarding the relevance of the above provisions to the interpretation of “tax advantage”, we reject any suggestion that they have any bearing on the analysis in this case. The utilisation of Speedy’s losses was a relief, and thus a tax advantage for the reasons already discussed. There are no facts suggesting any of the appellants or anyone else self-assessed on the basis of deemed pricing under the transfer pricing provisions or that there was any setting off of actual debits in Speedy or in the appellants due to the application of transfer pricing. Even if Speedy’s utilisation of the NTLRDs was considered to stem from the application of the transfer pricing regime, there would be still be a “relief” in the form of Speedy using losses and thus the necessary “tax advantage”. Similarly, it would make no difference to the analysis of the appellants’ debits if these were derived by reference to the deemed amounts generated through transfer pricing. The application of the debits would still amount to a relief.

The other strand to the appellants’ argument, that the FTT erred in concluding each of the appellants had an unallowable purpose, focusses on the FTT’s conclusion that the appellants’ purpose was to obtain a tax advantage for themselves, namely deductible debits.

The first aspect of this argument is that there could not be a tax advantage for the person incurring a debit where the person’s position vis-à-vis HMRC remained the same. This argument is rooted in the particular circumstances of the appellants in relation to which it is observed that that three out of the five appellants were loss-making before incurring the relevant debits and on the lack of any finding on the part of the FTT that any of the appellants’ respective charges to tax were reduced by the reorganisation. (The appellants accept that there are circumstances where a deductible debit is a tax advantage). We can address the appellants’ case briefly. The criticism is answered by the reasoning above (at [46] onwards) that for the purposes of 1139(2)(a), it is not necessary to determine whether the relief (here the function of a debit in reducing the amount on which tax may be charged) ultimately does result in less tax ultimately being payable, but just whether the provision operates in that way generally.

The appellants similarly argue the FTT erred in holding s441 applied to deny tax relief for the interest debits on the New Loans (to KF Finance and Stapleton’s). It is argued those appellants could not have had the purpose of obtaining a tax advantage either for themselves or Speedy because there was no finding by the FTT that any of the appellants had their respective charges to tax reduced by reason of the reorganisation for any of the relevant accounting periods. We reject this argument for the reason above.

As regards whether a company could have a purpose to secure the tax advantage of a deductible debit if it was loss-making, as Ms Wilson pointed out, a company can still have a purpose to do something even if the purpose is not ultimately successful. Moreover, even if a company were loss making, that would not preclude it having a purpose of reducing its own profits on the basis those could arise in future years or else against another group company’s profits by way of group relief.

We turn then to the second aspect of the argument. Here the appellant submits there was no evidence upon which the FTT could have reached the requisite finding that the appellantshadthe purpose (as distinct from knowledge) of increasing their own deductions. It is consequently argued that, in order to have found as it did, the FTT must have wrongly confused merely knowing that a debit would be deducted with having a purpose to secure that such debits would be deducted.

The principles relevant to the ascertainment of a person’s purpose in the context of unallowable purpose regime were most recently set out by this tribunal in BlackRock by reference to earlier authority (at [127]-[128], [168-169], and [177]). The decision in BlackRock was issued after the FTT decision, however the FTT was able to and did refer to the earlier authorities in its decision. The propositions were summarised in Ms Wilson’s skeleton, which the appellants did not dispute, and we gratefully set out those which are relevant here with some minor adaptations. In ascertaining the company’s main purpose:

Of particular significance, given its centrality to the appellants’ ground here, is the uncontroversial proposition, that knowing of a tax consequence does not equate to having a tax avoidance purpose. Thus, the appellants argue it is a non-sequitur to say that the knowledge that interest debits were deductible meant the borrower necessarily had a main purpose of obtaining a tax advantage.

In Versteegh ([160]) the FTT put the point this way in relation to the predecessor legislation contained in paragraph 13:

The FTT earlier emphasised in that paragraph:

The appellants emphasise that HMRC never put to the appellants’ witnesses that the appellants had a purpose of securing tax deductions for themselves. Mr Ghosh clarified the point was not so much regarding the unfairness of the FTT finding as it did, but rather because it meant that, given the lack of any other evidence, there was no material, which went beyond mere knowledge of the tax deduction, upon which the FTT could have made its finding that the appellants had such purpose.

HMRC draw attention to the need to look at the totality of the evidence, and with realism regarding the inter-linked nature of the two tax advantage elements of the restructuring: the increased debits for Speedy and deductions for the appellant debtor companies below Speedy. These amounted to a de facto surrender of losses down the group, something that was not permitted by the legislation at the time (i.e. the legislation that resulted in Speedy’s losses being regarded as “trapped”). Thus, in his e-mail of 20 June 2013 to the appellants’ solicitors Mr Andrews put the rationale behind the restructure as being that:

HMRC say there was ample evidence for the FTT’s finding regarding the appellants’ respective unallowable purposes. This took the form of the documentary evidence, admissions made by the appellants in the context of hearings (to do with closure notice applications) which preceded the FTT substantive hearing, and in the cross-examination of the appellants’ witnesses. There was no confusion regarding knowledge and purpose on the FTT’s part.

As regards the points made by the parties in relation to the evidence, we start with the two pieces of advice prepared for the appellants in relation to restructuring (the June 2013 Memorandum and the August 2013 Paper), which, in the appellants’ submission, point away from the appellants having the requisite purpose.

In relation to the June 2013 memo, Mr Ghosh referred to an illustration given in the section on “Tax Impact” (see [26] above):

In his submission, it was significant thatno-one had bothered to work out what the reduced tax charge would be in the borrowers.

The further August 2013 paper set out the following bullet points in its section on “UK corporation tax implications of the final position”:

Again, in Mr Ghosh’s submission it was significant that there was no evidence before the FTT that the detailed analysis referred to had been carried out.

Given the appellants’ case is then that the FTT lacked any evidence to make its finding regarding purpose, it is convenient to first outline the evidence which HMRC rely on for the finding on purpose, in order to then understand the appellants’ arguments that such evidence is insufficient to respond to the appellants’ ground.

HMRC took us to a number of pieces of evidence emanating from the oral cross-examination of the witnesses:

HMRC also referred to the appellants’ admissions in the context of their submissions in support of the appellants’ closure notice applications (which were before the FTT) that:

Mr Ghosh’s riposte is that these points all go the appellants’ knowledge of the effect of the restructuring. None of them address the lack of evidence for the FTT’s finding regarding the appellants’ purpose to secure deductions for themselves.

It is important, in our view, not to lose sight of the established propositions in the authorities that all the facts need to be considered in the ascertainment of a person’s object or purpose. The authorities also make clear the question is one of fact. The House of Lords in Brebner (Lord Upjohn at 30G) described the question a one for the first instance tribunal “to decide, upon a consideration of all the relevant evidence before them and the proper inferences to be drawn from that evidence.” The reference, both to considering all the evidence, and the acknowledgement that inferences would need to be drawn, recognises that questions regarding a person’s purpose are unlikely to be resolved by direct evidence or stated intentions alone.

The FTT had recourse to a great deal of documentary evidence and had the opportunity to hear, at length, live evidence from the appellant companies’ two witnesses. At [96] to [104], it made a number of detailed findings of fact, which it described under the heading of “findings of fact potentially relevant to purpose”.

In this context, to argue, as the appellants do, that the FTT’s finding that the appellants had a main purpose to secure tax deductions for themselves was made with no evidence is an uphill struggle. The appellants’ route to scaling those heights relies on the proposition that knowledge of tax effect is insufficient to establish a tax avoidance purpose.

However, this reliance overlooks that facts regarding the relevant company’s knowledge, while not by themselves conclusive, may still form a relevant part of the factual background. The appellants were clear they did not disagree with that as a matter of principle. The argument also effectively assumes a finding of fact concerning purpose will have its source in direct evidence when, as discussed above, it will be derived from a consideration of all the evidence and the drawing of inference.

The appellants also argued that it was not open to HMRC, given the way they had pleaded its case, to describe the tax advantage in the appeal before us as they had in their skeleton (by reference to generating increased tax relief for the borrowers, or to generating group relief) or to make oral submissions regarding an objective of creating flexible losses. This distracts, in our view, from the focus of our determination. That is not in how HMRC put their response before us regarding tax advantages. HMRC support the upholding of the FTT’s decision for the reasons it gave. Rather, what is relevant is the tax advantages which the FTT identified (use of NTLRDs by Speedy and deductions by appellants) and whether the appellants can persuade us that the FTT was unable to make its finding of fact on purpose concerning the appellants’ deduction of interest because of a lack of evidence.

In our judgment, it was clearly open to the FTT, on the basis of the evidence before it, to find the appellants had a purpose to secure deductions for themselves. That conclusion was encompassed within the accepted purpose of “using/accessing trapped losses” when taking account of the underlying motivation to the restructuring of achieving a group tax saving.

The appellants seek to draw the contrast between their accepted purpose of utilising or accessing Speedy’s losses on the one hand, with their disavowal of any purpose on their part in relation to seeking deductions for themselves, on the other. The fundamental difficulty with this is that it flies in the face of the clear rationale for the reorganisation - namely to achieve a group tax benefit. In the context of this restructuring, it makes no sense to talk of the appellants having a purpose of “using” Speedy’s losses in the abstract. The appellants’ submissions accept that simply increasing the speed at which the losses were utilised by increasing the interest rate would not save any tax (it simply accelerated an eventual charge to tax). The losses are only used, in any meaningful sense, if there is a profit somewhere in contemplation of which the losses can be set off against. In the context of this group restructuring, where profit was created in one entity by increasing income in Speedy, by moving intra-group debt, there would not have been any real “use”, or “accessing”, of the losses, if all that was done was to create more debt to pay in the paying entities elsewhere in the group. For the loss to be used in the sense it was intended – in other words in order to achieve a group tax saving - the debit generated in the paying entity had to then be capable of being put to use, by being set off against income, either in the paying entity or elsewhere in the group.

The legislation, as discussed above, is put in terms of the perspective of a particular company’s purpose and whether there was a purpose to secure a tax advantage for the appellants and/or for other persons. The individual company perspective did not mean the reality that the planning was approached at a group level was to be ignored in considering what purposes were at play from a given company’s perspective, just that the FTT had to be satisfied the group benefit purpose could be articulated in terms of a purpose in respect of the appellants themselves and/or other persons.

HMRC were therefore right, in the context of this case, to deal with the advantage to Speedy and the advantage to the appellants together, and to have put those related advantages to the witnesses as they did. That the appellants’ witnesses took no issue with the explanation put to them of the twin elements of the reorganisation could not have come as any surprise given the rationale for accessing the trapped losses in the first place.

The focus of Mr Ghosh’s point was that the witnesses were not squarely cross-examined on whether each of the appellants had a purpose to secure deductions resulting in an absence of evidence on the point. However, we consider there was no unfairness in the issue not being dissected in that way. In essence, the appellants’ case was that their purpose was to use or access the trapped losses. HMRC put to them that, in the context of the group level planning, that entailed the appellants being able to use the deductions. The witnesses had sufficient opportunity to explain why that was incorrect, but in fact confirmed the appellants’ deductions were a key feature of the reorganisation. The witnesses did not, in terms, give evidence that the appellants’ purpose was to secure deductions. But that evidence was not necessary. The FTT was well able to infer from the facts surrounding the reorganisation that there was such a purpose.

We agree with HMRC that neither of the points the appellants make about the advice memos undermine the FTT’s conclusion. Regarding the June 2013 Memorandum, the first bullet point (see [26]) was drafted on the assumption that the deductible debits created in the appellants would be capable of use. Regarding the August 2013 paper, if anything turned on whether there had been a detailed analysis, then it was for the appellants to adduce evidence on the point. The restructuring had gone ahead, the inference being that group relief issues were not an obstacle. Contrary to the reliance the appellants seek to draw from the point on the memos, what emerges is entirely consistent with the FTT’s findings: obtaining deductions in the appellants was a key element in the rationale for the reorganisation.

The FTT’s findings included, at FTT [98], that the appellants would not have gone ahead if HMRC had said the appellants would not get their deduction, Speedy theirs, or if there had been other adverse tax consequences (see [32]). We reject Mr Ghosh’s submission that it was not open to the FTT to take from this that the companies would not have done the transaction unless both elements (generating income in Speedy 1 – and getting deductions) were approved by HMRC. In other words, it was not possible to read this as saying the appellants’ deductions were material because it could have been the Speedy deductions, or other tax consequences which were material to going ahead. The straightforward reading of FTT [98] is that any of those factors would have stopped the transaction. The cross-examination on this point (see [96(4)]) also put the deductions on the part of the appellants as a central feature. There was no indication that the answer given was incorrect, and we see no issue with the FTT finding as it did regarding the appellants having a purpose that encompassed the appellants’ deductions.

The interlinkage of the two elements (utilisation of the NTLRDs, and deductions for the appellants) also explains why there is no illogicality, as the appellants argue, in HMRC’s stance (with which the FTT agreed) of capping the disallowance of debits to the amount of NTLRDs. While deduction of interest continues to be a tax advantage because deductions are a relief from tax, in accounting periods post-dating the exhaustion of the NTLRDs, it can no longer be said the appellants’ purpose in being party to the loan relationship was to secure those tax advantages. There is no dispute with the legal proposition that a company’s purposes may change over time.

On the facts of this case, there is also no difficulty, contrary to the appellants’ argument, with the FTT having considered the question of whether there was a main unallowable purpose with both tax advantages in mind. The appellants argue that as matter of principle each tax advantage should be considered independently so that its importance can be assessed (i.e. so it can be determined whether it is a “main purpose”) and so as to make the attribution test workable. However, as HMRC point out, the reference to “advantage” in s442(5), can clearly accommodate plural advantages by virtue of s6 of the Interpretation Act 1978, and where, as here, the tax advantages are linked, we consider there is no conceptual problem with assessing a company’s purpose in respect of both.

The fear advanced by the appellants, that analysing purpose in this way somehow allows purposes which are not by themselves “main” to be categorised as “main” when combined together begs the question by assuming that a purpose which encompasses two interlinked advantages should be considered singly. On the contrary, it would seem to run counter to the thrust of the scheme of legislation to exclude tax avoidance purposes which related to different tax advantages, which when considered cumulatively were “main” on the basis that individually they were not “main”. The clear dichotomy apparent from the structure of the legislation is that between tax avoidance purposes, on the one hand, and commercial purposes on the other, with less concern over distinctions between different types of purpose within those two categories. We also consider the wording of the attribution provisions, and the requirement that the apportionment is “just and reasonable” is flexible enough to accommodate analysis of multiple tax advantages. It would not preclude a tribunal apportioning by reference to a particular tax advantage (for instance if there was a different analysis regarding attribution for each). On the facts of this case, given the dependency between the tax advantages, it would appear to have been open in any case for the FTT to find each of the tax advantages were main purposes.

The appellants also take issue with the FTT’s finding regarding the significance of the interest rate not being increased on other loans, in particular the Detailagent loan. The FTT noted at [115(3)] that if Speedy had paid higher interest on that loan it would have had the effect of slowing down the utilisation of the NTLRDs. It viewed the omission to increase interest on the loan as strong evidence that the appellants acquired a new purpose when deciding to agree to increased interest charged on the pre-existing and New Loans owed to Speedy ([116]). Mr Ghosh submits the FTT’s reliance on the omission to raise interest was misplaced in view of Mr Andrews’ explanation to the FTT’s question to him on this point. Mr Andrews had suggested there would have been no point to the increase as it would be tax neutral with the payment and receivable cancelling each other out (Mr Ghosh had put it in terms that the debit to Speedy would be increased but the credit to ETEL would be increased too).

We see no error in the FTT drawing the inference it did from the appellants’ selective approach to the loans in relation to which the interest rate was increased. The FTT plainly was entitled to, and did, evaluate the explanation given for the omission of an increase in the context of the wider evidence regarding what the group sought to achieve through the reorganisation. An increase in interest rate would inevitably increase the debits and credits respectively of the debtor and creditor. The same would be true of the loans in respect of which interest was in fact increased. The FTT was right to remark on the selective nature of the rate increase because increased credits to Speedy were consistent with using up Speedy’s losses, whereas increasing debits were not. Describing an interest rate increase as being “tax neutral” for loans where Speedy was the debtor in essence reflected that an increase in that context would not use up losses. That stood in contrast to increases to interest on loans where Speedy was the creditor which would use them up.

We therefore see no error in the FTT’s finding of fact that each of the appellants had a subjective purpose to secure deductions for themselves.

The appellants also seek to argue the finding is not possible in view of the transfer pricing legislation. They argue that imposition of an arm’s length interest rate, in line with the transfer pricing legislation, to a loan with an underlying commercial purpose cannot intelligibly be described a giving rise to a tax avoidance purpose. However, there were no findings that the appellant companies were driven to become parties to loans at increased rates of interest in order to comply with such legislation, nor evidence which suggests the FTT was bound to find that that transfer pricing compliance featured in the appellants’ subjective purpose.

In the event HMRC succeeds in establishing that the appellants had a main tax advantage (which they have done given our conclusions above) the next issue is whether the FTT erred in its attribution of the debits to the unallowable purpose.

The relevant part of the legislation is set out in s441(3) CTA 2009 which stops the company bringing into account:

The Court of Appeal in Fidex Ltd v HMRC ([2016] EWCA Civ 385 – Kitchin LJ at ([74])) summarised the analogous predecessor provision in the following terms:

The relevant legal principles were set out by this tribunal in BlackRock (in addition to the Court of Appeal’s decision in Fidex the Upper Tribunal in BlackRock also had the benefit of that Court of Appeal’s decision in Travel Documents Service):

Ms Wilson correctly pointed out that the above should not be read as a new gloss to the statutory provisions and that the reference to the debits’ existence reflected the particular facts in BlackRock which concerned the existence of an SPV that had been created for a particular purpose. There is accordingly nothing in the above formulation which means it would not apply where there was already an existing loan relationship but one in relation to which a new purpose was acquired.

We would add to the above that the reference to the apportionment being “just and reasonable” accommodates the possibility that there may be more than one just and reasonable apportionment. The terminology of just and reasonable explicitly recognises the apportionment will be matter of fact-sensitive evaluation which we consider would be susceptible to interference on Edwards v Bairstow type grounds.

In developing his oral submissions, Mr Ghosh gave the example of a borrower who entered into borrowing for commercial purposes which paid “results-dependent” interest under the misapprehension the interest was deductible (it would be non-deductible because it was tantamount to a dividend) and who then, realising their mistake, refinanced the loan to one with deductible interest. If it were assumed the refinancing had the unallowable purpose of securing the tax advantage that would not, in his submission, mean the interest would be attributable to the unallowable purpose. He suggested that the debit, if it was the same or less as the old debit, is the same as it would have been under the old loan, then it would be attributable to the commercial purpose of that original borrowing.

We did not derive assistance from this sort of hypothetical example. The legislative provisions on attribution, as explained in BlackRock, involve an objective consideration of all the relevant facts and circumstances. Much would depend on the specific detail of the facts. It could equally be said the interest debits post refinancing were attributable to the unallowable purpose (refinancing the loan with the specific purpose of ensuring the interest was deductible). The suggested outcome also assumes the legislation operates so as to support the unwinding of a misapprehension whereas sometimes misapprehensions have far reaching consequences and attempts to unwind them do not work.

As regards the further formulation suggested by the appellants of asking “how much bigger” the debits were by reason of the unallowable purpose, we see little benefit to adding yet further to the formulations already established especially given the concern expressed in the authorities above that the words of the provision should not be glossed. As Ms Wilson’s submissions pointed to, the potential difficulty in any case with this formulation, in the context a situation where there are potential mixed purposes, is that it already assumes some of the debit is attributable to the purposes other than unallowable purposes whereas the extent to which that is the case ought to be a product of the determination.

The FTT considered the parties’ respective legal submissions on the approach to be taken. Although it was not able to take account of the Upper Tribunal’s decision in BlackRock (that hearing had not taken place by the time of the FTT Decision), the FTT in essence ended up in the same place by focussing on the provisions of the legislation without any gloss but acknowledging that analysing the matter by asking “but for…” could be of assistance.

In respect of the New Loans, the FTT attributed all of the debits to the unallowable purpose holding that the only purpose of KF Finance and Stapleton’s in being party to those loans was to secure the intended tax advantage for themselves and Speedy ([133]).

It also, in agreement with HMRC, attributed all of the interest increase to the pre-existing KFG loan with Speedy (in other words HMRC’s position allowing the existing 0.74% interest was maintained by the FTT). Although the FTT considered KFG had mixed commercial and unallowable purposes for being a party to the loan, the FTT held that the only purpose for KFG agreeing to the interest increase was an unallowable purpose. The FTT regarded the increase in interest as integral to the reorganisation and as strong evidence of KFG having a new purpose in being party to the KFG loan. The FTT rejected the appellants’ submission that the higher rate of interest would have been payable in any event because of the application of the transfer pricing legislation. That argument did not apply on the facts with the FTT noting the transfer pricing adjustments had not been made to the loans where interest rates had not been increased.

For the pre-existing loans in respect of which there had been an assignment of the loan to Speedy as the new creditor, rather than disallowing all of the interest as HMRC had, the FTT disallowed only the amount of increase. It considered HMRC’s approach did not reflect the extent to which the appellants had a commercial purpose as well as an unallowable purpose (the appellants would have continued to pay interest on their commercial borrowings at lower rates even if they had not participated in the reorganisation.) The FTT considered the debits to the original interest were to be treated in the same way as the debits to original interest on the KFG loan.

In relation to their appeal, HMRC argue the existing interest on the KFG loan and on the others were not comparable. The FTT wrongly overlooked its own findings that the loans were assigned in order to achieve the unallowable purpose. The only reason Speedy could access loan debits was as a result of assignment of receivables.

For their part the appellants argue the FTT was wrong, to attribute any interest to the unallowable purpose.

We deal with the appellants’ appeal on this issue first. They argue the FTT was wrong to disallow the interest to the extent it did, because there was no finding that any particular company had its position improved by reason of the reorganisation.

This echoes the argument the appellants made above that the appellants were not securing any tax advantage. However, the issue of attribution is only relevant, where, as is the case here, we have rejected the appellants’ argument on Issue 1. Under Issue 1 we concluded the FTT’s finding regarding the presence of tax advantages was correct. The absence of a finding that any particular company had its position improved was not an obstacle to finding there was a tax advantage. It is accordingly not an obstacle to finding that some or all of the debits were attributable to the purpose of securing the tax advantage of using Speedy’s losses and providing a deduction to the appellants.

The appellants also emphasise that it remained the case that the appellants required ongoing debt-funding for their commercial activities and submit that as such they were party to the loan relationships for a main commercial purpose. The FTT’s decision acknowledged the commercial purpose in the relevant creditor becoming a party to the loan relationship (as did HMRC) in relation to the existing 0.74% interest on that KFG loan. The FTT also took account of that commercial purpose to the relevant appellant becoming a party to the loan relationship in relation to the other pre-existing loans where the creditor had changed. The appellants’ point therefore concerns the increase in interest on the pre-existing loans and all of the interest on the New Loans.

As regards the increase in interest, putting aside the transfer pricing legislation argument which we will come to, there is no basis to say this increased amount would have been paid in any event; it is plain it arose wholly from the unallowable purpose. As regards the New Loans, there was no commercial purpose to KF Finance and Stapleton’s becoming a party; they only became a party to these loans because of the unallowable purpose. There was therefore nothing apart from the unallowable purpose to which the debits could be attributed.

The appellants also argue the relevant interest debits would have arisen to the appellants under transfer pricing legislation in any event. This argument was also made to the FTT. It reached the conclusions we have described above in the light of its findings of fact and conclusions it reached as to the purposes of the appellants ([132]). It had discussed the transfer pricing argument when considering purposes – that transfer pricing was not a purpose is evident from its findings at [115] which (at [115](4)) make clear the FTT was not satisfied the transfer pricing approach was applied to the loans; accordingly no claim was made or able to be made by the debtors for a corresponding adjustment.

We consider the FTT was correct to reject the appellants’ argument based on transfer pricing. The question, as simply put by the Court of Appeal in Fidex was “whether and to what extent the debit was attributable to the unallowable purpose”. The establishment of what purposes were in contention required a factual consideration the subjective purposes of the appellants in becoming a party to the loan. There are no findings which suggest compliance with the transfer pricing legislation in fact featured among the appellants’ subjective purposes.

The exercise of attribution is rooted in an objective assessment of the facts and circumstances. We agree with Ms Wilson that it does not require one to speculate what the position might have been if the appellants had chosen to comply with the transfer pricing legislation. Even if it were somehow relevant to take account that TIOPA deemed the arm’s length rate of interest of LIBOR + 5% (because if the relevant TIOPA conditions were met, the application of that rate via the legislation was not something either the taxpayer or HMRC could opt out of) it was up to the appellants as “disadvantaged persons” whether they chose to make a claim under s174 TIOPA. The fact is that they did not. There is no reason to suppose that the objective assessment of the facts and circumstances required for attribution should make the assumption that such claims were made when they were not.

Nor is there anything in the regime to suggest that attribution to the actual subjective purpose held, is then ousted by considering how the transfer pricing legislation would apply. The s441/442 regime specifically recognises (under s446 – see [74/75] above) deemed TIOPA amounts (where this has been applied). But the deeming of those debits cannot alter what the subjective purposes were of the company being denied debits and thus the possible sources in contention for the attribution. Those subjective purposes will be a matter of fact.

The only issue here concerns the debits relating to the 0.74% interest on the pre-existing loans where the debtor remained the same but where (unlike the KFG loan) the creditor was changed, by an assignment, to Speedy.

HMRC emphasise that in contrast to the original interest costs on the KFG loan (which Speedy was already sheltering by using its losses), in order to achieve the unallowable purpose, the other pre-existing loans had to be assigned and the interest increased. HMRC say the FTT wrongly ignored the assignment stage, which it had elsewhere in its decision accepted was integral to accessing the losses and to achieving the unallowable purpose. The only reason (all of) the debits existed under the loan relationship (to which Speedy was a party as a result of the assignment to it) was the unallowable purpose.

HMRC acknowledge the legislation’s reference to an apportionment which is “just and reasonable” signals the matter is one of evaluative judgment. But they submit the logical inconsistency, of allowing the existing 0.74% interest when it was accepted the mischief of the reorganisation (accessing the trapped losses) for which the loans carrying that existing interest were assigned, is a clearly an error of law: it is a finding contrary to the facts.

The FTT explicitly acknowledged there was “some merit” in HMRC’s approach in disallowing all of the interest but went on to explain that it did not reflect the extent to which the appellants had a commercial purpose as well as an unallowable purpose. In the FTT’s view, the appellants would have continued to pay the 0.74% interest on those commercial borrowings even if they had not decided to participate in the reorganisation.

We consider it was open to the FTT not to disallow the debits relating to the 0.74%. The fact it had found the assignment of the loan to Speedy was an integral part of accessing the losses did not stand in the way of that. The explanation for that can be found in returning to the legislation (s 442(1)(a) CTA 2009). That focusses on the question of the purpose for which the company seeking the debit was a party to the loan. The context in which the unallowable purpose of accessing the trapped losses arose was that it was included within the purposes for which the appellant debtor was a party in the relevant accounting period.

In so far as the purpose for which the company was a party to the loan was partly for existing commercial reasons, it was open to the FTT to consider that that purpose did not change because there was, post assignment, a new creditor. In other words, the purposes for which the relevant appellant was a party to the loan, even post assignment to the new creditor party Speedy, were not limited to the unallowable purpose because the appellant was a party for other reasons, at least in relation to the 0.74% amount of interest. The debits in the appellant still existed for commercial purposes up to the 0.74% notwithstanding that the creditor was different. The same was true of the KFG loan where the company was a party, at least in part, for reasons arising from the pre-existing commercial borrowing purposes. In contrast, in relation to the New Loans, KF Finance and Stapleton’s were parties solely for the unallowable purpose.

Finally, the appellants also argue as an overarching point that the FTT decision produces an absurd and arbitrary outcome in that it results in a “double-hit” to the Kwik-Fit Group. The appellants’ debits in respect of an arm’s length of interest are denied. Validly incurred NTLRDs available to Speedy are lost. But as HMRC point out, Speedy was able to use its losses. The appellants can still get a deduction, once the trapped losses are used up.

The fact deductions in respect of certain interest paid to Speedy are denied, despite it being arm’s length, simply reflects that the regime disincentivises situations where the company has the requisite subjective purpose of securing a tax advantage in becoming a party to a loan relationship or in entering into related transactions by attaching unfavourable consequences to that.

The appeals by the Kwik-Fit Group companies and HMRC’s appeal are dismissed. The FTT’s decision is upheld.